5.1 Triangulation

Triangulation is generally taken to mean the use of more than one method of data collection within a single research project. It can be used either 'across method' or 'within method' (Casey and Murphy, 2009). Across method is where both quantitative and qualitative methods of data collection are used. Within method means, within the context of qualitative research, that a mixture of qualitative methods of data collection are used, and for quantitative research, that a mixture of quantitative methods are used.

As well as triangulation of methods, there are other ways of triangulating your research, such as:

Time triangulation, where the data can be collected at different points in time (cross-sectional) or over a prolonged period of time (longitudinal)

Space triangulation, where the research is conducted in different countries or cultures. Within the Open University, this could also apply to research conducted with online and face to face tutor groups.

Researcher triangulation, where two or more researchers independently collect data.

Theoretical triangulation, where the research draws upon alternative or competing theories.

It should be noted that this section on triangulation owes much to the excellent section on validity and reliability in Cohen et al's (2007)Research Methods in Education.

Benefits of triangulation

The value of triangulation is that it gives multiple perspectives to your research. This is important not just in ensuring that you get a fuller picture of what is happening, but is also vital in convincing other people of the value of your research.

In particular, if you use different methods of data collection, particularly contrasting ones, then the more confident you can be in your findings. So, if a questionnaire and an observation produce consistent findings, then you, and those reading the research, can be more confident of those findings than if you just used a questionnaire or an observation in your research.

When considering research methods, therefore, you need to consider which methods will provide a particular angle or perspective on what you are studying.

External data, such as TMA scores and retention rates, could provide evidence of student performance and engagment. More quantitative methods, such as questionnaires, could provide an overview of usage of a particular technology or opinion of a change. More qualititiave methods, such as interviews, focus groups or observations, could provide insight into how people perceive particular issues or respond to change.

Validity and reliability

Two related issues in research which are associated with triangulation are validity and reliability. Particularly for those more used to quantitative approaches to research, these can be very difficult concept to understand and apply within qualitative research.

Validity means that the researcher has done their best to minimise bias in collecting the data, analysing the data and reporting the findings. Basically , what the researcher is aiming for is that the findings are supported by the evidence and that the evidence is based on honest analysis of data which is collected in a credible way. Bias can never fully be eradicated but the goal of the researcher is to mimimise bias and therefore maximise validity.

There are, in fact, many different types of validity. In quantitative research, validity can be maximised through:

  • careful sampling (are the participants selected randomly or representative of a larger group?)
  • use of appropriate methods
  • appropriate statistical treatment of the data.

In qualtitative research, it can be maximised through:

  • sampling (can the participants be said to be representative of a larger group?)
  • use of triangulated methods
  • peer debriefing or analysis of the data
  • respondent validation (to check factual errors or to allow the participant to respond? 
     

Reliability is particularly associated with quantitative research, where it assumes that the instrumentation (methods), data and findings should be controllable, predictable, consistent and replicable (Cohen et al, 2007, p. 148). The purpose is that variation in sampling, data collection and analysis has to be minimised. The result is that is the data was collected again, in the same conditions, with the same sample, and analysed in the same way, the result would be the same.

In qualitative research, reliability, in the way it is used in quantitative research, is more difficult and even thought not to be appropriate as an expectation of that research. Afterall, how can the findings of research conducted with a single tutor group by a single researcher ever hope to replicated with another group or by another researcher? However, there are still ways in which qualitative research can be reliable, although terms such as trustworthiness and dependability are often used instead.

The first approach to reliability in qualitative research is whether what the researcher records as data is what actually occurs in the research setting.

The second approach is that the researcher strives to record multiple interpretations and meanings given to situations and events (Cohen et al, 2007, p. 148). As with validity, this can be achieved through triangulation, peer debriefing and analysis, and triangulation.

Guidance on using triangulated methods

As indicated above, Cohen et al's classic Research methods in Education has an excellent chapter on validity and reliability. The 6th edition (2007) is availiable online from the Open University library.

Although relating to the field of nursing raher than education, the article by Dympna Casey and Katy Murphy, Issues in using methodological triangulation in research, is an excellent overview of how triangulate methods can be used in practitioner research. Details of the article can be found in the key refernces. It is available electronically from the Open University library.

Examples of research which has used triangulation

Angela Bowen was a consultant in the PILS (Personalised Integrated Learning Support) CETL. She used triangulation in her small-scale practitoner research project, Where is the shift/tipping point from initial student progress (DD121/122) and the take off to independent learning, self assessment and progression?. The report of her research can be found on the following direct link:http://intranet.open.ac.uk/cetl/pilsintranet/pics/d90300.doc

Wendy Fisher, a COLMSCT Fellow, did a research project on provision of tutor feedback using tablet PCs. The project used both interviews and focus groups as methods and is disseminated here as a final project report, Do we engage the student in e-assessment by personalising lecturers’ feedback interventions?

Research using triangulated methods

Last modified: Tuesday, 4 Mar 2014, 16:34